Adversarial Training on Purification (AToP): Advancing Both Robustness and Generalization

arXiv:2401.16352v2 Announce Type: replace
Abstract: The deep neural networks are known to be vulnerable to well-designed adversarial attacks. The most successful defense technique based on adversarial training (AT) can achieve optimal robustness against particular attacks but cannot generalize well to unseen attacks. Another effective defense technique based on adversarial purification (AP) can enhance generalization but cannot achieve optimal robustness. Meanwhile, both methods share one common limitation on the degraded standard accuracy. To mitigate these issues, we propose a novel pipeline …

abstract adversarial adversarial attacks adversarial training arxiv attacks cs.ai cs.cv defense networks neural networks robustness training type vulnerable